Research On Warning Effectiveness Of Crossing Collision Prevention System Based On Driving Style

With the rapid development of economy and the steady improvement of living standards,various transportation activities are increasing.Thereafter,it puts forward higher requirements for traffic safety management Traditional safety management methods focus on identifying and revising traffic accident-prone areas which is lacked of overall situation and it will spend a high cost of transformation.Based on the intelligent vehicle-to-infrastructure cooperative system,the theory of active road traffic safety planning takes people-vehicle-road as a whole,and discusses the problem of traffic safety governance at the height of the system.However,due to the differences between driver's driving style which caused by driver's intrinsic personality characteristics may greatly affect the actual effect of the vehicle-to-infrastructure cooperative system(CCPS).Therefore,it's necessary to take into account the driver's driving style when setting up the warning mode of the vehicle-to-infrastructure cooperative CCPS.Firstly,this paper introduces the concept and development process of intelligent vehicle-to-infrastructure cooperative system,compares it with traditional traffic management methods,and expounds its application value in traffic safety management And relevant traffic psychology studies are enumerated to prove the impact of driving style on driver's collision risk.Secondly,in order to verify the warning effect of the vehicle-to-infrastructure cooperative CCPS prompt mode on drivers with different driving styles in the actual driving process,this paper introduces a field of test of vehicle-to-infrastructure cooperative CCPS which was conducted in Toyota,Japan.Based on the experimental data,the changes of collision risk of drivers when they passed through the non-signal controlled intersection after receiving the system prompt information are analyzed,and the specific performance of drivers with different driving style after receiving different system prompt information is clarified.In this paper," Time to Collision(TTC)" and the risk level are used as the indicators to measure the collision risk.The influence of drivers' social attributes,traffic environment attributes,system information prompt mode(including no prompt information,“information+alert" mode,"alarm only" mode)and driving style attributes on drivers' safety is analyzed.And negative binomial regression model which was commonly used in traffic accident analysis and the Prais-Winsten regression model with panel-corrected standard errors and AR(1)which was commonly used in sequence correlation are selected in this study.By comparing and analyzing the fitting results of the two models,the specific performance of eight driving styles in different system prompt modes is clearly defined.In addition,considering the increase of the global elderly population,this paper also makes a comparative analysis of the elderly drivers(aged?65)and non-elderly drivers(age<65),and provides further suggestions for the suitable groups of different information prompting modes.The results revealed that these two warning modes have a good auxiliary effect on drivers with passive mentality when driving and drivers with natural worrier,but will increase the risk of drivers with scrupulous driving tendency.In addition,a more effective system prompting mode should be explored for drivers with confidence in driving,drivers with impatient driving tendency and drivers with signal-oriented preliminary preparing driving.